Pers. 45; Ps. 640; Si id domi esset mihi, iam pollicerer. Si intus esset, evocarem. Bacch. 554-5, Ps. 1320. This last example is somewhat complicated, but evidently is an explanation of Pseudolus' present attitude, and hence falls under this heading. Verging toward the explanatory sense are Rud. 202-3, 552-3. Asin. 678 and Aul. 439-40 are explanatory, but may refer to the past. Asin. 860; 2. Inferential. Pol ni vera ista essent, numquam faceret ea quae nunc facit. Merc. 382-3; Res adhuc quidem hercle in tutost. nam hunc nescire sat scio De illa amica: quod si sciret, esset alia oratio. In two cases it is impossible to tell whether the reference is to the present or the past. Cas. 555-6. Verum autem altrovorsum quom eam mecum rationem puto, Tri. 115; Haec, si mi inimicus esset, credo haud crederet. Referring to the past are Amph. 525-6, Aul. 742, Poen. 691-2. Cf. Cas. 910. Perhaps the most interesting result of this examination of Plautus' use of the present and imperfect subjunctive is the bringing to light the fact that, in the use of the imperfect tense, about two-thirds of the cases are either explanatory or inferential, whether we deal with those only which refer to the present or include those also that refer to the past. It is impossible to divide into clear-cut classes the cases that use the present subjunctive, and say that so many are ideal and so many unreal. that could be done, it would be possible to determine the proportion of explanatory and inferential in the sum total of the unreal, and thus make a comparison with the proportion found to exist in the use of the imperfect subjunctive. Though this exact comparison cannot be made, still a survey of the field leaves a strong impression that the proportion for the present subjunctive is less If than that for the imperfect. If this be true, the reason is not far to seek. The present subjunctive had been and still was, to a large extent, the accepted speech-form for the unreal as well as the ideal conditional sentence. Accordingly, in cases where the speaker was not forced to a conscious choice of a conditional thought-form, or his audience would arrive at his essential thought whether his words suggested to them the ideal or the unreal thought-form, he naturally chose the old familiar speechform. In only three (possibly five) cases does the speaker, in a situation of this sort, make clear by the use of the imperfect subjunctive that he is thinking in the unreal form. With the explanatory and inferential uses the case is different. Here the very essence of the meaning consists in the sentence being understood as unreal; hence the desirability of a speech-form to make this clear-the imperfect subjunctive was such a speech-form, now coming into use and ready to hand. It would be little wonder if it should prove to be true that there was a tendency to take advantage of it in cases of this sort, though the old speech-form, spoken in the proper tone of voice, could make the meaning clear. Interesting in this connection as showing the tendency to use an unambiguous speech-form for these special unreal uses is Men. 195; Nam si amabas, iam oportebat nasum abreptum mordicus. This is inferential, tending to show that Erotium affection is only simulated. Similar, but referring perhaps to the past, are Ps. 286, Rud. 379-80. Another matter of interest is to determine to what extent Plautus has adopted the imperfect subjunctive as the speech-form of the present unreal conditional sentence. This can be done roughly by comparing the absolute number of explanatory and inferential cases which find expression in the present and imperfect respectively, making some allowance for the possibility that these types appear in somewhat larger proportion in the imperfect. In the present subjunctive 13 explanatory cases were found; to these must be added, in this comparison, 12 examples1 that use forms in - am or - ar. Of the inferential there are 2 (or 5) cases; to these 2 are to be added for the same reason as above (Merc. 489, Men. 640.) 1 Asin. 393, Curc. 58, Merc. 286, Mil. 878-80, 1284-6, Poen. 877 (noverim), 971, Ps. 377, Rud. 196-7, St. 508, Tri. 628-9, Truc. 299. In the imperfect subjunctive were found 8 (or 10) explanatory and 2 (or 4) inferential. Combining these, the present subjunctive shows 29 (or 32), and the imperfect 10 (or 14) cases. From this we may assume that the present subjunctive is used by Plautus for the present unreal conditional sentence three times, where the imperfect is used once. With such a foot-hold as this, doubtless the imperfect subjunctive made rapid progress in displacing the present. UNIVERSITY Of California. H. C. NUTTING. NOTES. A NOTE ON THE ACHAEMENIAN INSCRIPTION, Bh. I, §18, lines 86-87. Dr. Louis Gray, A. J. P. XXI, page 21, reports a suggestion of Professor Jackson that the reading uša-bārim, adopted by Weissbach and Bang, should be retained; and adds that Jackson would render the epithet as 'borne by oxen,' and would equate uša- with Sanskrit ukşan and Avestan uxšan. We should, however, expect to find in Old Persian *uxša- rather than uša-, in accordance with the established law that in Av. and OPers. an Indo-European is gives a š, and IE. q's gives a xš. (See for examples Brugmann, Grundriss, I' §§616, 641, 819). To explain away this difficulty is the purpose of this note. = = This may, I think, be done by a reference to Pischel's law1 concerning the representation of ks and qs in Prakrit, namely by cch and kkh respectively. Examples cited in his Grammar for ks are: M., AMg., J. M. chuha = Av. šuða Skt. kṣudha; AMg., JM., S'.' acchi Av. aši, Skt. akşi; and, for q's: AMg., JM. khira Av. xšira, Skt. kşira. Pischel, ibidem, §320, further points out that Prakrit cch occasionally corresponds to Avesta xš. (But this is apparently not in conformity with phonetic law). As an example he cites uccha- = Av. uhšan [that is uxšan]; but he adds that there is another form ukkha-, which is authorized by the Prakrit grammarian Märkandeya. This seems to me to explain the apparently anomalous equation, OPers. uša- Av. uxšan = Skt. ukşan. For just as uccha- goes back to *uks- so does OPers. uša- go back to the same group; and as ukkha- goes back to *ugs-, so also does Av. uxšan. Whether Skt. ukşan goes back to *uks- or to *ugs- is impossible to 1GGA., 1881, p. 1322. Doubts are expressed by Johansson, Shāhbāzgarhi, II 20 ff. See, however, Pischel's Gram, der Prākrit Sprachen, §§319, 318 et passim. We have IE. kp not ks in acchi etc., but this is not important as IE. kp and is fell together in Aryan. See Brugmann, loc. cit. say; both IE. ks and qs alike become Skt. ks. We must therefore assume a "variation” of ễ and q, and that the "variation" was of Indo-European date. This occurs frequently. See Brugmann, Grundriss, I2 §597, 2, and the literature there cited; and also Wackernagel, Ai. Gr. I, page 228. Probably here IE. & was the more original, and the form with apparent q came as a loanword from the “centum-branch" to the "satǝm-branch." I close with the remark that Prof. Jackson's interpretation of the passage in question appears to be entirely justified. HARVARD COLLEGE, CAMBRidge, Mass., Feb. 16, 1901. TRUMAN MICHELSON. NOTES ON THE SEPTUAGINT TEXT OF II Sam. 7: 22 and Isa. 42: 21. II Sam. 7: 22. For ἕνεκεν τοῦ μεγαλῦναί σε (Cod. B) or ἕνεκεν τοῦ μεγαλυνθῆναι σε (Cod. A) read ἕνεκεν τούτου ἐμεγαλύνθης. Wherefore thou art great, O Lord.' The Septuagint rendering for 12 is usually dià Touro, but in Gen. 2: 24, 20: 6, and Hos. 136, it is evekev TOÚTOV. For the Hebrew original of ¿vekey toû, one would expect (Ex. 20: 20; II Sam. 14: 20), a (II Sam. 18: 18), or ¡ with infinitive (I Sam. 17: 28†; Ezek. 40: 4). It should be noted that with a noun, 'for thy word's sake,' appears in the preceding verse. It is therefore barely possible that the translator's eye may have rested on av in vs. 21 when he wrote eveкev Toû in vs. 22; it is more probable, however, that if this be the true reading, the Hebrew text was different from what we have now, perhaps y On the whole it seems better to correct the Septuagint from the Hebrew. The present reading can then be accounted for as follows. The original accurate rendering of the present Hebrew text, ἕνεκεν τούτου ἐμεγαλύνθης, was transmitted until a careless scribe wrote τοῦ for τούτου. Then someone, possibly the scribe himself, in order to provide eveкey with an object, changed the indicative to an infinitive, and added σe. Cf. I Sam. 26: 4 and Ps. 91: 6 (Heb. 92: 6) for other cases of éμeyaλúveŋv for or |